Beyond Text: Ethical Considerations for Visual Online Platforms
Author(s): Hannah Taylor
Hannah Taylor is a doctoral candidate in Rhetorics, Communication, and Information Design Program at Clemson University. Her research focuses on reproductive rhetoric, feminist methodologies, and digital rhetoric. Her work has been published in College English and Peitho.
Abstract: This article considers ethical decision making and privacy for a visual social media landscape and poses important questions for Internet researchers to consider before and during the research process. This author models self-reflective research practices by looking back on their own research practices in regard to two image-based social media projects: the online conference Braving Body Shame and the sexual health education Instagram page The Vulva Gallery.agency, embodiment, visual rhetoric
This conversation cluster comes out of two exigencies, both concerned with changes in internet research. As scholars of digital research ethics note, things change quickly in the digital sphere, which requires the field at large and individual researchers to enact reflexivity on our practices. Second, much of the current research focuses on text-based social media platforms, like Facebook and Twitter, that are easily scraped and collected at a mass scale. With the lessening popularity of Facebook and Twitter and the rise of TikTok and Instagram, which are far more image-based than their predecessors, researchers need to revisit their practices to be better attuned to the popularity of image-based social media. This does not mean reinventing the ethical considerations but thinking about them perhaps on a smaller scale for more image and video-based platforms. As the 2019 Association of Internet Researchers guidelines note, there is no universal research ethic for online research, particularly where privacy is concerned:
given the range of possible ethical decision-making procedures (utilitarianism, deontology, feminist ethics, etc.), the multiple interpretations and applications of these procedures to specific cases, and their refraction through culturally diverse emphases and values across the globe – the issues raised by Internet research are ethical problems precisely because they evoke more than one ethically defensible response to a specific dilemma or problem. Ambiguity, uncertainty, and disagreement are inevitable. (AoIR)
In keeping with the inevitability of ambiguity, I follow Katrin Tiidenberg when she suggests that “instead, an approach that reimagines ethical decision-making as a deliberative process that enables enacting beneficence, justice, and respect for persons on a case-by-case basis is increasingly recommended” (6). Like Leysia Palen and Paul Dourish, I am conceptualizing privacy as a boundary regulation process when they write that “privacy is not about setting rules and enforcing them; rather, it is the continual management of boundaries between different spheres of action and degrees of disclosure within those spheres” (3). Because of the increased disclosure that comes along with posting images online, the implications for privacy, sharing, and researching are heightened. This article considers ethical decision making and privacy for a visual social media landscape. It is important for scholars to ask: How are privacy and identity conceived of differently on primarily visual social media sites? What do internet researchers need to consider differently on visual platforms? What is at stake with embodiment in internet research?
In the following section, I discuss some of the previous questions that scholars have engaged in regarding research online before suggesting a framework for the ethics of online privacy in an increasingly visual social media landscape. Namely, I propose a series of questions that scholars can ask themselves before and during the research process in order to address ethical and values-aligned research studies utilizing visual data. As people put more and more of themselves online, it is important for scholars to continue revising our tactics of protecting both our identities and the privacy of our participants. To do this, I discuss my research experiences with two image-based social media projects: the online conference Braving Body Shame and the sexual health education Instagram page The Vulva Gallery. Finally, I reflect on my own research practices in order to demonstrate a feminist research ethic of self-reflexivity.
Approaches to Privacy and a Heuristic for Image-Based Platforms
The dynamic world of social media research requires scholars of many backgrounds to think critically about their research practices. This conversation contribution is less concerned with what happens in cases of automated data collection and more about the situations where the identity of the research subject is the focus of the research. Scholars across disciplines use internet research to look at language and social action. As Amber Buck and Devon Ralston note, writing studies professionals use internet research to discuss student writing practices, digital rhetorical practices, participate in digital ethnographies, and more. It is important within all of these contexts to consider “who is conducting that research, how communities are represented, and how that data is collected and distributed are key concerns for writing researchers and point to the need for considering critical digital literacy in research ethics” (Buck and Ralston 3). Buck and Ralston provide an excellent heuristic for reflective research and data collection. I build on this heuristic by providing further considerations for video and image-based media, while also demonstrating a process of reflection that is necessary for ethical engagement with social media research. Although using someone’s words in research carries risk, attaching a face and body to that information is even riskier, particularly with the rise in deepfakes and identity fraud.1 Scholars have begun to discuss how to protect the privacy of individuals when their identities are easily searchable—such as when their name or likeness is contained in a post—and IRB has measures to address this risk. IRB also attends to voice and image included through interviews and videos. But now, scholars must update our approach to address voice and image in online data collection.
Buck and Ralston’s exploration of key considerations is the jumping off point for this discussion, which formulates four questions for scholars to consider: 1) Who owns the post? 2) What is the network of the user? 3) How is the user engaging with their own privacy? and 4) What are my research values?
Buck and Ralston thus encourage us to question who owns the posts, where they can be shared, and how the presence of a researcher complicates those two questions. These questions are further complicated depending on the positionality of the social media users. Stephanie Vie explains that “it’s critical to consider as researchers whether and how to share and recirculate those stories, particularly when they’re being shared by individuals from vulnerable populations” (262). Lauren Cagle speaks to researchers’ positionality and asks scholars to consider their agency and engage with participants about… where participants’ information falls along the “public/private continuum’’(7). In other words, scholars have to consider their role and consider the implications of further sharing posts from participants whose consent was not given.
The public/private continuum is complicated on social media, where images often travel beyond intended audiences. Social media platforms often collapse audiences, “making it difficult for people to use the same techniques online that they do to handle multiplicity in face-to-face conversation” (Marwick and boyd, “I Tweet,” 120). As a result, even if a poster is sharing private information, they may not have the intention of a broad, public audience. This can be especially true on sites like Instagram and TikTok where popular “Explore” pages and “For You” feeds send content to a wide range of people. These types of platforms ask users to have a less defined audience, making ethical research more complicated. Researchers need to consider the ethical implications of shifting audiences to an academic space by including posts in their data set. When people post images or videos online, it is highly unlikely that they imagine a group of academics discussing their posts. For example, if a person posts a political rant online for their friends and family, they may not envision that a researcher of political rhetoric would engage with them, nor present their work to another group of researchers. In cases where informed consent is not possible, such as one of the case studies I discuss later on, it is key that researchers consider what is at stake for posters and their identities. Therefore, asking Who owns the post? Is the owner different than the original poster? What are the implications of sharing this image in a different context than its intended space? can lead to greater contemplation and ethical engagement with online data. It is true of both text and image posted to social media that users’ posts often travel beyond their intended audience. With images, and the potential implications of likeness being shared, researchers need to consider ethics beyond the original poster’s desire and broaden thinking to focus on networks.
One way to complicate engagement with visual posts is to consider the role of networked publics in digital research. danah boyd defines networked publics as “spaces constructed through networked technologies and imagined communities that emerge as a result of the intersection of people, technology, and practice” (Marwick and boyd, “Networked Privacy,” 1059). Alice Marwick and danah boyd take up this concept and note that viewing privacy beyond a binary will allow for a community-based approach to information sharing, rather than an individual one. In this conception of privacy, networks are negotiated between changing audiences, “social norms, and technical affordances” (Marwick and boyd, “Networked Privacy,” 1064). To illustrate this concept in action, consider Michael Zimmer’s discussion of a research team looking at Facebook accounts of college-students. Despite the steps taken by the team to protect the users’ privacy, the identities of those in the database were easily discovered. Zimmer contends that had the researchers adopted a more networked view of online information, they could have better shielded the participants from discovery. This view of privacy allows scholars to further nuance their approach to information published online. In some cases, as Zimmer points out, users set specific restraints to limit who sees through the social media platform itself. Some platforms, like Tik Tok, are designed to have a more expansive networked reach. In the case of Zimmer, lack of attention to the power of networks allowed for the identities of individual participants to be easily found. Contrastingly, Alice Marwick and danah boyd looked at the privacy of teenagers and their posting online. They found that teenagers saw privacy as a matter of boundaries, primarily consisting of trust and context. They argue that “conceptualizing privacy as networked highlights the difficulty involved in defining or even understanding social contexts, as they are co-constructed by all present and shaped by the affordances of the social technology in play” (“Networked Privacy,” 1063). The key difference between these approaches is that Marwick and boyd centered the teenagers’ understanding of privacy and their knowledge of their own networks. Therefore, they were able to approach the participants based on social contexts that the teens were already aware of and comfortable in. Additionally, considering networks has the potential to protect research participants because it makes researchers more aware of the many ways that privacy can be breached across a variety of contexts. Therefore, scholars should ask themselves: Who does the users’ network include? How is the network potentially impacted by the platform they are using?
The emphasis on network also raises the question of identity. Often, researchers will come to a specific online community because of the identities of the communities they are studying, so they must be mindful of the cultural implications of online engagement. Privacy is culturally situated. For example, Catherine Knight Steele discusses how Black communities may share public information online but intend for it to remain within that discourse community like a type of oral cultural community (116). Sharing social media information outside of its intended socio-cultural situations can put marginalized communities at risk of harm. As Zimmer notes, “merely having one’s personal information stripped from the intended sphere of the social networking profile and amassed into a database for external review becomes an affront to the subjects’ human dignity and their ability to control the flow of their personal information” (321). Therefore, it is important to view each post in terms of the users’ cultural experience and ask: Am I interacting with a culture that might define privacy differently than I do? This question may not have an obvious or clear answer but demonstrates that researchers should engage in some reflexivity about the assumptions we bring to the question of privacy.
Beyond these more subject-focused questions, an ethical approach to research should acknowledge the role of the researcher. How am I defining my own ethical engagement? How are my values as a researcher reflected in my process of researching? For example, as a feminist researcher, it is important for me to center the lived experiences of my subjects. I subscribe to a feminist ethics of care that is both “a value and a practice, both affective and cognitive” (Tiidenberg 7). It requires researchers to be mindful of power dynamics and ask sticky questions of representation. This approach to research ethics necessitates a reflective process where individuals can confirm that I am interpreting their experiences and intentions correctly. In the absence of this possibility, it is necessary for my analysis to be careful and supported. The work of feminist research is not simple–there is emotional labor present in care-based ethics, and an approach that prioritizes individual autonomy and experience is not always the most effective for a research project that aims to be more generalizable. It is important, however, for researchers to establish their individual value of ethics beyond the pragmatic concerns of IRB.
I began this conversation by noting that the landscape of social media is constantly changing. At our current cultural moment, it is difficult to say what online research will look like in a few years’ time. Therefore, any approach to internet research ethics needs to be flexible and self-reflexive. In the following section, I detail my experiences working with two datasets shared in visual formats and the ethical considerations I undertook while doing this. I use my work not because it is exemplary, but because I believe it is important that researchers are transparent in their practices, even when we might make different choices in the present. It is this amount of self-reflexivity that will lead to ethical engagements with internet research.
Case Studies: Braving Body Shame and The Vulva Gallery
To demonstrate these guiding questions in action, I will discuss two research projects, one finished and one ongoing, that helped me shape this approach to researching social media images online. The Braving Body Shame conference first took place in the spring of 2020 and featured speakers from a variety of backgrounds. The conference was virtual and took place over a week. I was initially drawn to the conference because of the explicit focus on embodiment and shame, and I analyzed the various ways that participants in the conference described their experiences of overcoming shame. I examined nine video interviews from actresses, activists, dancers, and students who discussed their feelings of shame and how they have worked to move past it (Taylor). This study initiated my interest in internet research ethics as I had to consider my use of these videos for the purpose of academic publication.
The second case study I discuss involves the Instagram page, The Vulva Gallery. The page features illustrations of people’s vulvas with their stories of embodiment and acceptance. Though the posts do not feature individual’s faces, they often include their name, information about their family, friends, and locations, and unique experiences that could threaten anonymity. Both of these research sites were places where participant experiences were already grouped together on an online platform, so I knew that the individuals had agreed to have their information shared beyond their immediate followers or network. This does not, of course, as discussed above, assume that the individuals imagined that their materials would one day be the center of research studies.
Who Owns the Post?
In the case of Braving Body Shame, the conference owned the posts. In my analysis, I did not link to the speakers’ platforms outside of the conference, aiming to honor their wishes about where and how their information is shared. However, my ethical considerations did not end there. Despite the fact that the conference was open, the participants in the Braving Body Shame conference mentioned explicitly that their content was not aimed at academic audiences. In fact, the organizers state the misconceptions in academic research as the exigency for their entire conference. During my writing and review process, this tension came up fairly frequently. The home page of the conference still states:
After attending a couple of in-person academic conferences, one of our hosts saw that there was a BIG part of knowledge and understanding missing from each conference. She realized that there was a great NEED for a conference that was more accessible and less academic-focused. (Braving Body Shame).
The conference organizers felt that academic discussions of body shame often removed the lived experience of individuals, favoring generalizable and quantitatively driven information. How did I, then, as a researcher, justify researching a group whose stated exigency was to move away from academic audiences? More importantly, how would I protect their likeness as I worked to analyze it? First, I ensured that my research goals aligned with the conference goals – to focus on the lived affective and embodied experience of people experiencing body shame. Second, the speakers encouraged viewers to share the information widely, without any caveats about academic research. I was never able to reach the conference organizers after trying several times throughout the research process. Third, I only studied video interviews that had been shared beyond the conference (see explanation in next section).
With the Vulva Gallery, I had a different experience. It was much clearer how to protect the identity of participants, partially because the posts were already anonymous. The images were illustrations, therefore protecting the likeness of the individuals, and I could protect their privacy by following the example of the page. The Instagram page and gallery owned and posted the image, and therefore the reach was broad. The participants submitted their own images for the purpose of education and empowerment. For example, the About page on the website states that “The Vulva Gallery is an online gallery and educational platform celebrating vulva diversity, aiming to improve sexual health education and opening up conversation about topics that are still being stigmatised.” The educational purpose behind the postings reveals that the audience is intentionally broad. The participants agreed to have their images shared via a popular social media site and are aware that the audience they may be reaching is larger than their individual follower-base. Beyond considering ownership, looking at the networked publics of the posts allowed for more nuanced analysis of ethical considerations.
What Is the Network of the User?
Beyond the specific audience of the individuals, it is important to consider the broader networks that they engage, specifically with how their posts are distributed based on platform affordances. In the case of Braving Body Shame, the audience was not markedly different from the network of the participants, at least at first. The videos were originally posted on a website for conference purposes only and required a password to access. After the conference, however, the videos were posted on YouTube with the consent of the participants, according to the conference organizers. This move made me feel more confident in my use of the data, as it was clear that the participants consented to their talks being shared beyond the initial audience of the conference. I initially received feedback from my article reviewers that I needed to more clearly justify why I was using this information at what seemed against the wishes of the conference organizers. I explained that each of the individual participants posted their videos. For analysis, I only drew from YouTube videos that had been highly circulated and suggested high public engagement. Some of the more popular participants had tens of thousands of followers and linked the videos to their Instagram feeds. Because these were public figures, the question of ethics was less about protecting their identities and more about considering the agency and decision-making of the participants. Had the videos only existed on the Braving Body Shame website, I do not think I would have proceeded with my research. Furthermore, the networked of speakers expanded far past conference attendees because the speakers employed the affordances of the platforms. Speakers shared clips of video interviews, spliced together parts of their talks, and reshared both to their feeds and stories on Instagram. Again, this intentional public reach gave me confidence in including these materials in online research.
Similarly, The Vulva Gallery used the affordances of Instagram to expand the network of people who see the vulva illustrations. My engagement with these posts was also about honoring the intention and agency of the participants. According to the owner of the site and illustrator, Hilde Atalanta, the participants submit their own images and stories to be published on both the website and Instagram. The individuals are not directly connected to the page through tagging, so their direct exposure to the network of the page is limited unless they comment directly on the post itself. The Instagram page had already done the work of considering consent and network, meaning at the very least that participants knew that their images would be shared to a public space. However, as other scholars note, this was not informed consent to participate in my research, begging the question of the benefits of doing this research and potential harms for participants. The Vulva Gallery fills a similar gap in popular sex education as it does in academic research—there are very limited discussions of diverse bodies in health education and related scholarship. Academic research on visual representations of female body parts, especially sexual organs, is primarily focused on harm. This research adds an empowerment focus. In addition, my research goals aligned with the purpose of the gallery–to introduce narratives of diverse bodies into sexual education. These factors gave me confidence that I was honoring participants’ intentions and not introducing more harm than they were already exposed to.
How is the User Defining Their Own Privacy?
Beyond an analysis of the network that the information was shared in, it is important to consider how the individuals consider their own privacy. In both cases I discuss, I was studying diverse populations that were united by a common identity or experience. This, however, did not ensure that each individual person considered privacy in the same way. Within the context of Braving Body Shame, the participants were part of marginalized and multi-marginalized communities. Each speaker experienced some level of discrimination based on their size, and many experienced oppressions related to their race, gender, ability status, or sexuality. Though participants primarily discussed body shame, this affective experience was never fully separate from their other experiences of shame. Each of the speakers noted that they wanted to share their experiences of body shame so that others’ experiencing shame and people perpetuating shame can learn from their experiences. They consented to the videos being posted for conference participants and then for the public at large. I did not engage with any material about the subjects outside of the video interviews in an attempt to maintain the amount of privacy that they agreed to.
Following my experience with Braving Body Shame, I closely considered the potential harm and how the participants considered their own privacy for Vulva Gallery participants. The Vulva Gallery is a pseudo-anonymous site; most of the stories contain first names and identifiable information such as location and experiences that are unique. However, this information was presented on Instagram and the gallery’s website. I chose to not include any of the images of vulvas or stories that could be identifiable in my research, favoring quotes that contributed to a thematic analysis. The identities of the participants were hidden aside from what they chose to reveal in their narratives, making it difficult to discern if they thought about privacy differently based on cultural differences. Still, by consciously working to not introduce harm and considering individuals’ definitions of privacy, researchers can work toward more ethical online research.
How Am I Defining My Research Values?
As I described earlier, my approach to online research is grounded in feminist research ethics, and as I walked through the above questions before determining whether or not to finish and start these projects, I made it a priority to ground my analysis in the lived experiences of participants. I centered both the Braving Body Shame conference participants and Vulva Gallery posters’ goals in my work and withheld any impulse to critique, instead prioritizing how they publicly framed their personal experiences. As may be evident from my detailing of this decision-making process, the ethics felt clearer in the Vulva Gallery than Braving Body Shame, despite the fact that the former contains more private information. This clarity is in part because of the nature of the public information and partially because I refined my approach to ethical considerations.
These guiding questions are not a comprehensive list of things to consider, but they do provide a heuristic for examining the ethical implications of researching visual platforms. Because the posts are more embodied, the researcher should carefully consider the material consequences of their research. By asking these questions before engaging in research, scholars are more likely to treat participants with ethics and care.
Conclusion, or a Moment of Reflexivity
In an attempt to honor my feminist research ethic, the writing of this conversation contribution has made me re-evaluate my own orientation to participants’ privacy and the value of my work. Particularly, I was much more aware of my research ethics in analyzing the Vulva Gallery because of the questions posed by reviewers during the process of first publishing the piece on the Braving Body Shame conference. I did not receive formal training in online research, and so my initial question, analysis, and consideration left out the negotiation process that is privacy setting. This is perhaps an argument for more training on online research ethics in graduate school, and more broadly an example of a feminist research reflection that interrogates decision-making. I would have likely considered a different set of questions if I had been thinking about the agency of the participants beyond availability of the Body Shame video interviews. This reflective process will facilitate my ethical research decisions in the future, as well as model processes for interrogating the complicated relationship between participants’ privacy, agency, and networked engagement for other researchers.
As many researchers have expressed, wading through the constantly changing landscape of social media requires a re-evaluation of research processes. The set of questions I propose is just one example of the many ways that researchers can approach their ethical considerations, and I invite scholars to build upon this heuristic in establishing best practices for digital research on visual materials. Like Buck and Ralston, I acknowledge that “issues of privacy and surveillance are fraught and always changing on social media platforms” (10). The relationship between public information online and privacy concerns will continue to blur and following a feminist ethics of centering the experiences of individuals is one way to honor the complexities of the platforms and people we study.
“About.” The Vulva Gallery. https://www.thevulvagallery.com/about
“About.” Braving Body Shame Conference. https://www.bravingbodyshame.com/
Buck, Amber M., and Devon F. Ralston. “I Didn’t Sign Up for your Research Study: The Ethics of Using ‘Public’ Data.” Computers and Composition, vol. 61, 2021, pp. 1-13. https://doi.org/10.1016/j.compcom.2021.102655
Cagle, Lauren E. “The Ethics of Researching Unethical Images: A Story of Trying to Do Good Research Without Doing Bad Things. Computers and Composition vol. 61, 2021, pp. 1-14. doi: 102651.
franzke et al. Internet Research: Ethical Guidelines 3.0. 2020. https://aoir.org/reports/ethics3.pdf
Marwick, Alice, and boyd, danah. “Networked Privacy: How Teenagers Negotiate Context in Social Media.” New Media and Society, vol. 16, no. 7, 2014, pp. 1051–1067
— “I Tweet Honestly, I Tweet Passionately: Twitter Users, Context Collapse, and the Imagined Audience.” New Media & Society, vol. 13, no.1, 2011, 114–133. https://doi-org.libproxy.clemson.edu/10.1177/1461444810365313
Palen, Leysia, and Paul Dourish. “Unpacking” Privacy” for a Networked World.” Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. 2003, pp. 129-136.
Steele, Catherine Knight. “Black Bloggers and their Varied Publics: The Everyday Politics of Black Discourse Online.” Television & New Media, vol. 19, no. 2, 2018, pp. 112-127 https://doi.org/l0.1177/1527476417709535.
Taylor, Hannah. “Unsticking Shame: Considering Lived Experience and Processes of Overcoming.” Peitho, vol. 25, no. 1, 2022. https://cfshrc.org/article/unsticking-shame-considering-lived-experience-and-processes-of-overcoming/
Tiidenberg, Katrin. “Research Ethics, Vulnerability, and Trust on the Internet.” Second International Handbook of Internet Research, edited by Jeremy Hunsinger, Matthew Allen, and Lisbeth Klastrup. Springer, 2020, pp.569-583. https://doi.org/10.1007/978-94-024-1555-1_55
Zimmer, Michael. “‘But the Data Is Already Public’’: on the Ethics of Research in Facebook.” Ethics Information Technology, vol. 12, 2010, pp. 313–325. DOI 10.1007/s10676-010-9227-5